基于高分二号卫星影像高潜水位煤矿区沉陷地土壤含水量监测

Monitoring of soil moisture in coal mining subsidence area with high-level groundwater based on the GF-2 satellite image

高潜水位地区煤炭开采破坏导致地表沉陷出现积水和斜坡,沉陷内土壤含水量会分布不均匀,影响农作物的生长,从而严重影响矿区居民的生产和生活。因此大范围快速、精确监测高潜水位地区煤矿开采区的土壤含水量具有重要现实意义。卫星遥感技术可以快速、准确、高效监测矿区土壤含水量。通过遥感手段对高潜水位采煤塌陷地土壤含水量进行监测,探求出一个比较方便、快速、合理监测高潜水位采煤塌陷地土壤含水量分布状况方法,为矿区环境影响评价、农作物估产、破坏等级评价、耕地损害补偿与土地复垦方案的编制提供参考依据。借鉴土壤含水量遥感监测经验,通过野外实地采集土壤样本并测量土壤光谱数据,在室内测量土壤含水量,分析实测地面光谱数据与土壤含水量的变化关系,结合实测的土壤含水量与光谱特征数据,对土壤含水量与实测水体光谱进行相关性分析,得到土壤含水量光谱数据敏感波段范围。结合高分二号卫星影像谱段数据特点,将实测光谱波长按照波段范围划分为与高分二号卫星影像谱段对应的4个波段,即450～520,520～590,630～690,770～890 nm,再取各个波段范围反射率的平均值与土壤含水量光谱反射率进行相关性分析,寻求高分二号卫星影像监测土壤含水量最敏感的波段数据,在确定遥感探测敏感波段的基础上,建立了土壤含水量与光谱反射率的遥感反演模型,即:S曲线模型、逆函数模型,基于预处理的高分二号卫星影像进行沉陷区地土壤含水量遥感反演,从而得到高潜水位采煤塌陷地土壤含水量的空间分布情况。研究结果表明不同土壤含水量的光谱特征基本相似,实测地面光谱数据与土壤含水量的变化关系为土壤光谱反射率随着波长的增长而增大,呈正相关关系;土壤含水量与高分二号卫星影像数据B3波段的反射率具有显著的负相关关系,可将B3波段作为监测土壤含水量最敏感的波段;通过对S曲线模型、逆函数模型进行分析与检验,S曲线模型比逆函数模型更接近实测值;基于高分二号遥感影像,利用S曲线模型进行遥感反演,可以迅速得到高潜水位采煤塌陷地土壤含水量空间等级分布图。

Coal mining could destroy and lead to surface subsidence resulting in water accumulation and land sloping in the areas with a high submersible water level.The water content in the subsided soil will not be evenly distributed,which seriously affects the growth of crops,and the life of residents in the mining area.Therefore,it is of practical significance to monitor the soil water content in the coal mining area with a high submersible water level. Satellite RS technology can quickly and accurately monitor the soil moisture content of mining areas.The RS technology is used to monitor the soil water content in the coal mining area with a high submersible water level,and to find a more convenient,rapid and reasonable method for monitoring the distribution of soil water content in the coal mining area with a high submersible water level.It provides a reference for the environmental impact assess-ment of mining areas,crop yield estimation,damage grade evaluation,farmland damage compensation and the compilation of land reclamation plan.Based on the experience of RS monitoring of soil moisture,the soil samples in the field are collected,the soil spectral data are measured and the relationship between the measured ground spectral data and soil moisture is analyzed by indoor soil moisture measurement.Combining the measured soil water content and spectral characteristic data,the correlation between soil water content and measured water spectrum is analyzed to obtain the sensitive band range of soil water content spectral data. According to the characteristics of GF-2 image band data,the measured spectral wavelength is divided into four bands corresponding to the image band of GF-2,namely 450-520,520-590,630-690 and 770-890 nm.Then the correlation between the average reflectance of each band and the spectral reflectance of soil moisture content is analyzed.The most sensitive band data for monitoring soil water content with GF-2 is found.On the basis of determining the sensitive band of RS detection,the RS inversion models of soil water content and spectral reflectance are established,which are S curve model and inverse function model.Then the RS inversion of soil water content in subsidence area is carried out through the pre-processed GF-2,and the spatial distribution of soil water content in coal mining subsidence area with a high groundwater level is obtained.The results show that the spectral characteristics of different soil moisture content are basically similar,and the relationship between measured surface spectral data and soil moisture content demonstrates that the soil spectral reflectance increases with the increase of wavelength and the soil moisture content has a significant negative correlation with the reflectance of GF-2 image data in B3.Therefore,the B3 band can be used as the most sensitive band for monitoring soil water content.Through the analysis and test of S-curve model and inverse function model,the results indicate that S-curve model is closer to the measured value than inverse function model.Based on the RS image of GF-2,using S-curve model for RS inversion,the spatial grade distribution map of soil water content in coal mining subsidence area with a high groundwater level can be quickly obtained.